Population-based diagnostics pinpoint the power in numbers

 

The world of diagnostics is expanding. Instead of focusing on individual animals, population-based diagnostics help veterinarians and producers identify the health status of their barns more quickly and efficiently. Jeff Zimmerman, DVM, PhD, a professor at Iowa State University, is helping pave the way for this trend. A decade ago, individual pig samples accounted for 52% of all samples. Now, that number is about 22%. In addition, processing fluids and oral fluids are increasingly important tools for detecting disease.

 

“The population diagnostics we were using in the 1980s and 1990s don’t apply very well in our larger populations,” Zimmerman told Pig Health Today. “By that I mean, as we get bigger systems with more buildings and more pens, randomly sampling 30 pigs doesn’t work anymore.”

 

 

The use of oral fluids and processing fluids has allowed the tests to be adapted for aggregate samples, he explained.

 

“We can have very good detection rates, with very good diagnostic sensitivity and specificity,” he said. “Our producers can do better surveillance with less work. For the industry today, we need good surveillance so people understand what infectious agents are circulating in their populations and they can respond correctly.”

 

 

“The phenomenon of low-prevalence PRRS [porcine reproductive and respiratory syndrome] is something we hadn’t really anticipated, but it’s becoming a widely recognized fact,” Zimmerman said, explaining that “low-prevalence” refers to the virus existing at a level of 5% or less in herds. “When you start collecting individual samples to detect that level of prevalence, it takes a lot of testing.”

 

The question, however, becomes where to draw the line on getting an accurate picture of the disease profile in a building by using processing fluids or oral fluids.

 

“We’re getting better estimates on how many negatives you can pool in one positive — or add to one positive — before you start losing detection,” he said. “Part of what drives [that question] is how important the correct answer is to the producer. In other words, what level of false-negatives can you actually tolerate in your system?

 

“That’s one part of the equation,” he added. “The other part of the equation is how frequently do you sample? Those are the dynamics we’re working with now, and I think the industry is finding its way to the right balance.”

 

 

Zimmerman said population diagnostics for PRRS is a priority because it is helping researchers improve testing algorithms and assays, but the protocol is being used for many other diseases, including foreign animal diseases like African swine fever.

 

“We can use molecular technology to create vaccines that will produce the antibody response we’re looking for,” Zimmerman explained. “For example, there’s an antigen called P30 in African swine fever that is important in diagnostics. We can actually create a vaccine that will just produce a P30 antibody, so we don’t have to work with a pathogen.”

 

 

There’s no other way for the US pork industry to be prepared for foreign animal diseases except by having oral-fluid-based assays ready to go, Zimmerman said. “We don’t have enough people in the field to collect all the samples we’d have to collect.”

 

At the first positive detection of a disease, “we need to be ready to sample lots of herds, but we also need an aggregate sample that represents dozens of individual pig-bleeding samples,” he continued.

 

The assays belong to the USDA and are in that department’s purview, so the industry can’t run assays for African swine fever, foot- and mouth disease or classical swine fever “just because we want to,” Zimmerman said. “But we can help USDA get ready [with test development].”

 

The industry also can assist with surveillance and in determining how to collect samples in a strategic fashion that will best detect the pathogen.

 

 

The Pseudorabies Eradication Project involved random sampling, and the classic number was 30 randomly selected pigs that were bled and tested, Zimmerman said. “We know that 30 is no longer the right number…and we’ve learned that pathogens move spatially.”

 

For example, if one pen in a barn is infected, the pens immediately adjacent to that pen will likely become infected next, and the virus will move in this fixed spatial pattern throughout the barn.

 

“The assumption for random sampling is that the target is random. We know that’s no longer the case, so we need to revamp and reevaluate our sampling strategy,” Zimmerman said. He wants to continue testing to understand how this applies at a site, county or state level.

 

“There is power in numbers, but you still need the right numbers, and you have to interpret them correctly,” he said.